Physics-informed neural network estimation of active material properties in time-dependent cardiac biomechanical models

2025-05-06

AI Summary

Clinical cardiac imaging provides only displacement and strain data, lacking direct measurements of active myocardial stress—a critical limitation for biomechanical assessment. Method: We propose a physics-informed neural network (PINN) framework to inversely estimate spatiotemporally varying active stress parameters in a time-dependent myocardial biomechanical model. Our approach innovatively integrates adaptive weighted loss, Fourier feature embedding, structural regularization, and a dual-network architecture decoupling state and parameter fields. We further provide the first systematic analysis of how loss weighting influences active parameter reconstruction accuracy. Results: The method achieves high-resolution, robust active stress field reconstruction across multiple noise levels, significantly enhancing spatial detail fidelity. It enables quantitative identification of myocardial heterogeneity and fibrotic scar regions, delivering an interpretable, clinically translatable, noninvasive tool for quantitative assessment of cardiac fibrosis.

Technology Category

• Machine Learning: Active Learning
• Planning, Routing, and Scheduling: Activity and Plan Recognition
• Computer Vision: Medical and Biological Imaging

Application Category

• User Modeling, Personalization and Recommendation: Psychology-informed user models and recommender systems
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Active stress models in cardiac biomechanics account for the mechanical deformation caused by muscle activity, thus providing a link between the electrophysiological and mechanical properties of the tissue. The accurate assessment of active stress parameters is fundamental for a precise understanding of myocardial function but remains difficult to achieve in a clinical setting, especially when only displacement and strain data from medical imaging modalities are available. This work investigates, through an in-silico study, the application of physics-informed neural networks (PINNs) for inferring active contractility parameters in time-dependent cardiac biomechanical models from these types of imaging data. In particular, by parametrising the sought state and parameter field with two neural networks, respectively, and formulating an energy minimisation problem to search for the optimal network parameters, we are able to reconstruct in various settings active stress fields in the presence of noise and with a high spatial resolution. To this end, we also advance the vanilla PINN learning algorithm with the use of adaptive weighting schemes, ad-hoc regularisation strategies, Fourier features, and suitable network architectures. In addition, we thoroughly analyse the influence of the loss weights in the reconstruction of active stress parameters. Finally, we apply the method to the characterisation of tissue inhomogeneities and detection of fibrotic scars in myocardial tissue. This approach opens a new pathway to significantly improve the diagnosis, treatment planning, and management of heart conditions associated with cardiac fibrosis.

Problem

• Estimating active cardiac material properties from imaging data
• Improving active stress parameter reconstruction with PINNs
• Detecting myocardial tissue inhomogeneities and fibrotic scars

Innovation

• Physics-informed neural networks estimate cardiac material properties
• Adaptive weighting and Fourier features enhance neural network accuracy
• Energy minimization reconstructs active stress fields with high resolution